Pinned Tweet
Agents are your new users. Learn from them.
AX Cloud is now in research preview: managed infra, agents and data for experimenting with coding agents as users.
Run unbiased simulations of coding agents at scale. This allows you to understand how agents discover and use products. Yours, your partners' and your competition's. Use this to optimize your product, craft go-to-market materials or develop your technical partnerships.
Get started for free on managed infrastructure, then upgrade when you're ready to parallelize at massive scale. Every run starts from a fresh sandbox. No bias. No context pollution. No cheating.
Run your first experiment in minutes: docs.514.ax/getting-started
“The very first experiment we did with 514 turned up gaps in our CLI that were tripping up agents as they tried to build with Mux. After an immediate fix, we were able to reduce tool call failure by 34%.”
— Dylan Jhaveri, Director of Self Service at Mux
Run your first experiment: docs.514.ax/getting-started
79
Agents are your new users. Most teams still learn how they behave from a handful of chat sessions and a gut feel.
That’s anecdata. AX is what you get when you treat agent experience like something you can measure — same task, fresh sandboxes, scored outcomes, at scale.
82
Every AX Cloud run starts in a clean sandbox. No prior chat history. No leftover files. No way for the agent to peek at the tests.
That’s how you get an unbiased read on how coding agents actually discover and use your product.
514.ax
80
Fiveonefour retweeted
Today @vermilionfund closed our $25M second fund.
Still backing technical founders building for technical users in AI infra, security, and dev tools.
95% of my Fund I investors came back for Fund II.
That's the number I'm most proud of.
@DominicMadori from @TechCrunch wrote about our story
techcrunch.com/2026/07/08/so…
defineQueryModel(): one TypeScript metrics definition, typed to your ClickHouse tables. Built for agents. Strict types, local dev server, and LSP make dev-agent cycles fast. Metrics created reduce the surface area you expose to your runtime agents.
Wrote about the approach with Nakul Mishra from @AWS and Al at @ClickHouseDB.
Blog: clickhouse.com/blog/metrics-…
Demo: github.com/514-labs/financia…
Most AI database migrations fail.
Here's what actually works for Postgres → ClickHouse.
clickhouse.com/blog/ai-power…
What happens when fitness data actually drives revenue?
F45 Training saw a 70% CSAT lift and 50% lower costs! We’re breaking it down live in LA at the Connected Health & Fitness Conference.
Join us here: connectedhealthandfitness.co…
84
Watch how we use the open-source Moose Stack to embed real-time analytics and AI chat into a Next.js app—powered by ClickHouse and Red Panda streaming.
Here is the intro, but you can watch the full demo of the future of analytics in applications here:
youtube.com/watch?v=41eQikJc…
139
Tomorrow’s the day!
Join Chris, ClickHouse and F45 Training as we dive into how real-time analytics transformed their member experience.
📅 November 20 | 9 AM PT | 12 PM ET
🔗 clickhouse.com/company/event…
1
84
🚨 5 days until we go live with ClickHouse and F45 Training!
Real-time analytics, 1B+ data points, and a 70% boost in satisfaction — all in 30 minutes.
Save your seat 👇
📅 November 20 | 9 AM PT | 12 PM ET
🔗 clickhouse.com/company/event…
The magic behind real-time analytics? Partnership.
We helped F45 Training build a scalable data engine that delivers results in milliseconds, using ClickHouse and MooseStack.
Learn how on Nov 20 👇
🔗 clickhouse.com/company/event…
5
118
What does 1 billion data points really mean for a fitness brand?
For @F45Training, it’s real-time visibility into performance — across members, workouts, and studios.
Join us on Nov 20 to see how real-time data transformed their member experience.
🔗 clickhouse.com/company/event…
5
67
Can AI be your next data engineer?
We think yes—if you go dev-first.
Watch the full conversation → youtube.com/watch?v=6iPmENP7…
28
MooseStack native APIs:
If you want "OLAP in your App", you need to expose your #OLAP data to your App, usually with APIS. In today's part one, #MooseStack's native APIs (coming up, bring your own API framework). fiveonefour.com/blog/ship-yo…
38
Modern analytics shouldn’t leave you waiting: they should talk back.
Join our own Chris Crane for a free session on integrating realtime + agentic analytics into your app with MooseStack.
osacon.io/register/
Customer result: sub-second analytics replace 20–30s waits.
Recipe: AI + context (schema/queries) + guardrails (types/engines/sort-key policy).
Case study: Mike Klein modeled 100+ views in ~4 hours. Link in reply. #Analytics #DataEngineering
Moose + SQLModel = Python OLTP → OLAP, end to end.
We rebuilt our TypeScript demo in Python to show what really happens when your ORM meets ClickHouse.
Same schema, a few more manual seams than typescript.
Still way cleaner than the usual ETL spaghetti.
Read: fiveonefour.com/blog/just-ol…
Repo: github.com/514-labs/oltp-to-…
1
81
Fiveonefour retweeted
I built a code-first, local-first CDC pipeline to ClickHouse using MooseStack.
- spins up Postgres, Debezium, Redpanda, ClickHouse in 1 cmd
- consume Debezium topics in code
- apply stateless transforms
- manage ClickHouse tables in code as the sink
Repo: github.com/514-labs/debezium…





